Much of the basketball analytics is useless when it comes to utilising it to improve the team performance. That’s for a variety of reasons.
- Measured factors are not properly defined. For example, a contested shot has proven to be a difficult case.
- Factors are defined invalidly. For example, eFG% does not validly measure the efficiency of field goal shooting.
- Conclusions are invalid. For example, generally speaking, mid-range 2P shots are less efficient than close-range 2P and 3P shots. Yet you can’t conclude that one should not shoot any mid-range 2P shots.
- There’s a limited understanding about the transference from practice to game performance. Using analytics to find out where improvement is most urgent or most potential doesn’t do any good unless you understand how the mediation works in complex systems.
Now, this can be improved. We can better utilise performance analysis to enhance the game performance.
Take eFG%, for example. It can replaced as the go-to measure for the efficiency of field goal shooting.
My suggestion is a formula called Shooting Quality Assessment Tool or SQAT.
In SQAT, we calculate the point production of a single field goal shot considering not just if the shot goes in but also if it leads to a shooting foul or an offensive rebound. This is the equation for the point production of a single shot:
(2 x 2P + 3 x 3P)
+ (FTA x 0.75)
+ Shooting foul x 0.10
+ (ORB after FGA) x 0.99
= Points produced
So, if the shot goes in, it’s two or three points, depending.
For each FT earned, it’s 0.75 points. This number depends on the average FT% of the league where the game is played.
For each shooting foul, there’s a 0.10 point extra. That’s an estimated shared advantage due to the increased probability of a further foul during the quarter leading to at least one bonus FT set.
If there’s a miss & no shooting foul & an offensive rebound, the point production is 0.99 or the average point-per-possession production in the league.
Given this, the point production of different types of field goal shots are:
- 2P missed + DRB = .00
- 3P missed + DRB = .00
- 2P missed + ORB = .99
- 3P missed + ORB = .99
- 2P missed + shooting foul = 1.60
- 2P made = 2.00
- 3P missed + shooting foul = 2.35
- 2P made + shooting foul = 2.85
- 3P made = 3.00
- 3P made + shooting foul = 3.85
This formula is from my Master’s dissertation back in 2014. In my sample, close range 2P shots produced more points per FG shot than 3P shots did (1.42 and 1.30, respectively) even though the eFG% in both categories was the same (57%) (when there was no shooting foul).
However, close range 2PS led to a higher FT frequency and ORB%. This goes to show that SQAT really is different from eFG%.
Per SQAT, mid-range 2P shots were the least productive (0.93 points per FGS). Their eFG% was 38%, and only seldom did they lead to FTA or ORB.
If you’re interested in the details, see the whole dissertation here.
SQAT could be improved. For example, the formula doesn’t take into account that the FT% rises as the set continues. So, the second FT of a set is more likely to drop than the first, and the third one more likely than the second one.